The Impact of a Construction Land Linkage Policy on the Urban–Rural Income Gap

Abstract

Promoting coordinated urban–rural development represents a key policy initiative by the Chinese government to advance rural revitalization and promote common prosperity. As a central component of China’s land management system, the Urban–Rural Construction Land Linkage Policy aims at dismantling the historical urban–rural division while fostering balanced regional growth. This research analyzes panel data spanning 2010–2022 across 294 prefecture-level cities, utilizing a multi-phase difference-in-differences (DID) approach to evaluate the policy’s effectiveness in reducing urban–rural income disparities. Empirical findings reveal that the policy implementation has substantially narrowed the income gap between urban and rural populations. Heterogeneity analysis indicates that the policy’s impact is more pronounced in China’s eastern regions. Mechanism analysis reveals that the policy narrows the income gap through two primary pathways: first, by promoting urbanization through facilitating rural-to-urban population transfer and optimizing urban spatial layout. Second, by driving industrial structure optimization through intensive land use that advances agricultural scale and modernization, while improved land resource allocation boosts secondary and tertiary industries. These findings offer empirical support and policy insights for refining urban–rural land management strategies and advancing integrated development.

1. Introduction

Globally, the unequal distribution of land systems and resources between urban and rural regions has contributed to an expanding development gap [1]. In response, developed countries often rely on market-oriented mechanisms (such as transferable development rights in the United States) or planning strategies (like the compact city policy in the European Union) to mitigate these disparities. In China, the urban–rural divide remains a prominent challenge. According to the National Bureau of Statistics, in 2023 [2], the per capita disposable income of urban residents was 51,821 CNY, compared to 21,691 CNY for rural residents, resulting in an urban–rural income ratio of approximately 2.39:1. This economic gap is closely linked to institutional disparities in land use rights and governance structures between urban and rural areas. China’s distinctive dual land system presents institutional obstacles that severely hinder integrated urban–rural development [3]. Since the reform and opening, the pace of urbanization in China has quickened, accompanied by issues such as the abandonment of rural homesteads, underutilization of industrial and mining land, and the emergence of numerous “hollow villages”. Simultaneously, urban land use has been marked by disorderly expansion and low efficiency, further intensifying the mismatch between land supply and demand and obstructing balanced urban–rural growth [4,5].

Against this backdrop, China has implemented more stringent farmland protection policies and institutional measures to promote intensive land use, shifting the focus of land supply from expansion to stock optimization. China has begun to implement a strict farmland protection policy and a system to promote intensive land use, shifting the focus of land supply from quantitative expansion to inventory optimization [6]. The Urban–Rural Construction Land Linkage Policy (also known as the land quota trading policy) aims to foster urban–rural integration by optimizing land resource allocation. Under this policy, rural idle and inefficient construction land can be reclaimed as farmland, while the resulting land-use quotas may be transferred to urban development, thereby maintaining a balance in total urban–rural construction land and optimizing land-use structures [7].

The Urban–Rural Construction Land Linkage Policy differs significantly from the land management systems adopted in countries such as those in Europe, the United States, and Japan. In Europe and the United States, transferable development rights (TDR) enable the spatial reallocation of land development rights through market-based mechanisms [8,9]. The TDR mechanism is market-oriented, emphasizing property rights protection and flexible regulation, and is implemented under the guidance of local governments to balance ecological conservation with development needs [10]. Japan’s “Land Readjustment” system has been implemented since the early 20th century. It is characterized by a government-led cooperative mechanism that optimizes land use and controls costs on the basis of guaranteeing the right to original residence and fair compensation, thus effectively avoiding the social conflicts that often arise in large-scale land acquisition and relocation [11]. The Community Land Trust (CLT) system in the United Kingdom operates through community-based nonprofit organizations that retain land ownership while providing long-term leases to ensure housing access and community stability for low-income populations. Through this approach, CLTs create institutional safeguards for both housing security and urban–rural integration [12]. In contrast, China’s Construction Land Linkage Policy features a dual-characteristic system that combines government-led planning with market-oriented mechanisms. This approach not only emphasizes planned quota allocation and cross-regional land balance under administrative guidance but has also incorporated market-based operations in selected pilot areas. The policy’s operational framework addresses both surface-level resource allocation optimization and deeper structural challenges, including the growing urban–rural income disparity, imbalanced factor distribution, and increasing pressures on agricultural land conversion [13,14,15].

Compared with previous studies, the marginal contributions of this research are threefold: First, this study addresses a critical gap in the literature by systematically examining the urban–rural income gap and its underlying mechanisms under the Urban–Rural Construction Land Linkage Policy. Second, while existing studies present conflicting findings regarding the policy’s effectiveness, our research extends the discourse by investigating the policy’s indirect impacts through two key transmission channels: urbanization progression and industrial structure transformation and their subsequent effects on income distribution. Third, departing from previous case-specific analyses that focused on individual regions, this study provides the first systematic evaluation of the Linkage Policy’s aggregate effects while accounting for regional heterogeneity. Our findings not only supplement the limited evidence on the policy’s nationwide implementation outcomes but also reveal significant interregional variations in policy effectiveness.

The research framework of this paper is as follows: Section 2 presents the literature review, Section 3 presents the theoretical analysis and research hypotheses, Section 4 presents the materials and methods, Section 5 presents the results, and Section 6 presents the conclusions and recommendations.

2. Literature Review

With reference to existing studies, this paper will systematize the evolutionary logic and implementation effects of the Linkage Policy.

2.1. Evolutionary Logic of the Linkage Policy

China’s Construction Land Linkage Policy represents an institutional innovation designed to regulate urban–rural land relations within the country’s unique political–economic framework [16]. In 2005, the General Office of the State Council of China issued the Circular on Promoting the Pilot Work of Linking the Increase and Reduction of Urban and Rural Construction Land, marking the official entry of the policy into the institutionalization of the pilot phase [17]. In its initial stage, the policy relied primarily on administrative quota allocation, serving two main purposes: maintaining the “one-to-one compensation” balance between occupied and replenished arable land and rehabilitating idle rural construction land. This was achieved by converting vacated rural homesteads and construction land back into arable land, thereby generating quotas for new urban construction land development [18].

With the implementation of China’s New Urbanization Strategy, the policy has evolved beyond its original scope, gradually incorporating multidimensional objectives such as increasing farmers’ income, facilitating rural revitalization, and enhancing market-based allocation of land resources. This transformation has elevated the policy from a singular land management instrument to an integrated framework that actively contributes to national spatial governance and promotes urban–rural coordinated development [19]. The evolution of the policy’s fundamental logic demonstrates China’s innovative governance approach to reconciling the urban–rural land system dichotomy through restructuring development rights and land use entitlements [20]. This institutional innovation operates through a dual mechanism: (1) establishing quantitative linkages between rural land rehabilitation and urban development quotas, while (2) creating an interregional quota trading platform. This integrated framework simultaneously addresses urban land scarcity while capitalizing on rural land assets through market-based valuation [21].

With the ongoing policy evolution, a significant institutional development occurred in 2020 when the CPC Central Committee and the State Council issued the Opinions on Constructing a More Improved Institutional Mechanism for the Market-based Allocation of Factors. This document proposed the principle to “explore establishing a nationwide inter-regional trading mechanism for construction land indicators and arable land compensation quotas” [22]. Subsequently, the 2022 Opinions on Accelerating the Construction of a Nationwide Unified Big Market further institutionalized this approach by mandating the integrated planning and unified management of both incremental and existing construction land resources [23]. The exploration of establishing a nationwide inter-regional trading mechanism for construction land quotas represents a crucial step in deepening market-oriented reforms for land factor allocation. This institutional innovation carries profound significance by creating systematic channels for equitable exchange and bidirectional flow of production factors between urban and rural areas while facilitating the regional redistribution of land value-added benefits [24].

In summary, the evolution of China’s land governance policy demonstrates a systematic transition from “centralized planning and control” toward “market-oriented regulation”, embodying the characteristic gradualism of China’s reform approach within the dual urban–rural land system framework. Moreover, the central government has actively encouraged local policy innovation, resulting in diverse implementation approaches including interregional land quota trading mechanisms, integrated urban–rural development zone initiatives, ecological compensation linkage systems, and marketization pathways for collectively-owned construction land [25]. While significantly enhancing policy effectiveness, this institutional development has simultaneously generated regional implementation disparities.

2.2. Implementation Effects of the Linkage Policy

Since its inception, China’s Construction Land Linkage Policy has emerged as a pivotal mechanism for advancing land system reforms and facilitating urban–rural spatial transformation. Academic research has systematically evaluated the policy’s impacts across four key dimensions: (1) enhanced land-use efficiency, (2) strengthened arable land preservation, (3) increased rural household incomes, and (4) optimized urban–rural spatial configurations. Empirical evidence demonstrates that the policy has achieved significant outcomes in resource utilization and spatial reorganization, including the reactivation of underutilized rural construction land, intensified utilization of urban construction land, and measurable improvements in both the quality composition and the allocation effectiveness of cultivated land resources [26,27]. In terms of improving the well-being of farmers, studies have shown that the Linkage Policy has contributed to the improvement in rural infrastructure and living conditions through land consolidation and revenue-sharing mechanisms and, to a certain extent, has increased farmers’ property income and welfare levels [28]. The policy’s economic benefits have been used to support rural industries, which has also led to an increase in the income level of farmers [29]. Through its land quota trading system, the policy addresses urban construction land shortages while simultaneously enabling interregional land resource mobility and spatial restructuring between urban and rural areas [30,31]. Case study evidence from pilot regions including Chongqing and Hubei demonstrates the policy’s effectiveness in enhancing land utilization efficiency, safeguarding agricultural land, boosting rural household earnings, fostering urban–rural connectivity, and promoting more balanced regional development [7,31].

However, the policy implementation has exposed several institutional shortcomings and governance challenges. On the one hand, some local governments, to pursue land financial gains, rely excessively on index trading, which has led to the phenomena of low quality “linkage” and “false reclamation”, resulting in the decline of the quality of land reclamation and the alienation of the policy objectives [32]. On the other hand, the policy implementation suffers from opaque rights distribution mechanisms, resulting in insufficient farmer participation in land value appreciation benefits. This imbalance has created an asymmetric benefit structure where local governments and property developers capture many gains from land assetization [32]. In addition, the performance of the Linkage Policy varies from one region to another owing to the capacity of the Government to implement it as well as to regional differences [33].

In summary, the urban–rural Linkage Policy has demonstrated measurable success in enhancing land resource allocation and fostering urban–rural integration, while effectively contributing to spatial restructuring, farmer welfare improvement, and local fiscal expansion. Nevertheless, current research remains inadequate in providing a systematic empirical evaluation of the policy’s distributive effects, particularly regarding its mechanisms and impacts on urban–rural income disparities. This study consequently employs urban panel data and econometric modeling to comprehensively examine the policy’s influence, transmission channels, and regional variations in affecting the urban–rural income gap, with the objective of offering empirical support and policy insights for optimizing the linkage mechanism’s design and equity outcomes.

3. Theoretical Analysis and Research Hypotheses

The wide gap between urban and rural incomes is one of the major challenges facing China today. The income disparity between urban and rural residents not only reflects the imbalance in regional development and the imbalance in the distribution of social resources but also directly affects social equity, economic efficiency, and social stability. Among the many influencing factors, the imbalance in the allocation of land factors and the distortion of the mechanism for distributing land value-added gains in the process of de-farming are widely considered to be one of the core mechanisms causing farmers to remain in a long-term low-income trap [34].

According to the neoclassical convergence theory, the urban–rural income gap is considered a temporary phenomenon. The mobility and diffusion of production factors are expected to gradually eliminate regional disparities [35,36]. The Linkage Policy facilitates rural land capitalization and property rights establishment through land element reallocation, increasing farmers’ income and reducing urban–rural disparities. Regarding property income, it enables farmers to gain stable earnings through land leases and share dividends by advancing land rights formalization and transfer mechanisms, thus achieving land assetization [31,37]. In terms of operating income, the Linkage Policy improves the utilization efficiency of rural operating land through land integration and function optimization, promotes the optimization of rural industrial structure, and drives the increase in farmers’ income [28]. In terms of transfer income, the policy releases the value of land assets through the market mechanism and provides funds for local finance [38]. These revenues are used to invest in rural public services and compensate farmers, further improving their ability to benefit from fiscal transfers [37,38]. In terms of wage income, the policy promotes land improvement and infrastructure construction, creating temporary jobs and industrial employment opportunities during implementation and raising the employment rate and wage level of the rural labor force. Concurrently, by upgrading rural infrastructure and public service provision, the urban–rural construction land linkage narrows inter-sectoral service disparities, consequently mitigating urban–rural income inequality. Building upon the extant literature, this investigation advances the following research hypotheses:

Hypothesis 1.

The Urban–Rural Construction Land Linkage Policy contributes to reducing income disparities between urban and rural populations.

Industrial structure, as a core driver of China’s rapid economic growth, profoundly influences the dynamic evolution of the urban–rural income gap. The upgrading of the industrial structure has played a significant positive role in narrowing the urban–rural income gap [39]. The policy enhances urban–rural land allocation efficiency through systematic consolidation of underutilized rural construction land and strategic expansion of urban development space. In rural areas, the policy stimulates endogenous motivations for intensive land use through rational quota allocation and incentive mechanisms. Moreover, the newly developed land under the Land Linkage Policy supports various uses, including residential construction, infrastructure, and rural industrial development, facilitating structural transformation in rural areas [29]. The policy facilitates rural land revitalization by enhancing arable land consolidation and optimizing underutilized land efficiency, thereby ensuring land supply stability and creating spatial foundations for modern agricultural development [40]. In urban areas, the policy has released the potential of land appreciation and provided a space guarantee for urban industrial development [41]. On the one hand, land remains a crucial means for local governments to attract enterprise investment. The use of the Urban–Rural Construction Land Linkage Policy enhances industrial development, particularly in secondary sectors [42]. On the other hand, the additional land provides support for the development of residential land, and when the cost of land and labor is high in the eastern region, the relocation of labor-intensive industries will provide for the development of tertiary industries in the relocation area [43].

The optimization of industrial structures across urban and rural areas has fundamentally reshaped rural economic composition, facilitating the coordinated flow of key developmental elements—labor, capital, and land resources. This structural transformation has propelled rural regions into a new phase of transitional development, industrial upgrading, and comprehensive revitalization [14,44]. As a result, these dynamics, while significantly raising the income levels and livelihood sustainability of farmers, have also greatly facilitated urban–rural integration and reduced the urban–rural input gap. Building upon the extant literature, this investigation advances the following research hypotheses:

Hypothesis 2a.

Industrial upgrading serves as a mediating mechanism through which the Urban–Rural Construction Land Linkage Policy influences the urban–rural income gap.

As a critical production factor, land inevitably plays a key role in shaping new urbanization [45]. Increasing farmers’ incomes and economic growth relies heavily on higher urbanization rates [46]. The goal of urbanization is to improve quality of life through qualitative rather than quantitative changes, emphasizing industrial upgrading and the transformation of employment structures rather than mere population migration or land utilization [47]. Thus, urbanization and narrowing income inequality are inherently aligned [48].

The land development opportunities created by the Linkage Policy provide essential space for urban industrial growth and public infrastructure development while simultaneously enabling significant rural-to-urban labor migration that enhances cities’ employment absorption capacity [49]. Through facilitating organized population movement and industrial factor concentration, the policy contributes to transforming urbanization from quantity-driven expansion to quality-focused development. The rural-to-urban labor migration facilitated by the Linkage Policy directly boosts agricultural household wage earnings, elevating overall rural income levels. Concurrently, urbanization-induced industrial upgrading and service sector growth generate higher-value employment opportunities for both urban and rural populations, progressively reducing interregional income disparities [48]. Furthermore, the policy’s extension of urban infrastructure and public services to rural areas advances basic service equalization while strengthening rural development potential and social inclusion [40]. Collectively, these mechanisms establish urbanization as both an economic growth driver and an institutional framework for sustained urban–rural income convergence. Building upon the extant literature, this investigation advances the following research hypotheses:

Hypothesis 2b.

Urbanization rate serves as a mediating mechanism through which the Urban–Rural Construction Land Linkage Policy influences the urban–rural income gap.

4. Materials and Methods

4.1. Data Sources

This paper collects data related to 294 prefecture-level cities over the period 2010–2022 and uses a multi-period double-difference (multi-period DID) model to assess and analyze the impact of the policy on the urban–rural income gap. The data are mainly from the China Regional Economic Statistics Yearbook. It should be noted that the implementation of the urban–rural construction land increase/decrease Linkage Policy is carried out gradually in phases and regions, and the data sources are mainly based on the websites and trading platforms of the Ministry of Natural Resources (MNR) and provincial departments of natural resources, so the pilot data in this paper are mainly obtained through manual collection and organization. Figure 1 shows the cities where the policy was implemented and the specific years in which it was implemented. In this paper, the prefectures implementing the policy of linking urban and rural construction land are treated as a treatment group, and the prefectures not implementing the policy as a control group. Before constructing the model, the data were shrunk by 1% up and down, and finally 3822 valid samples were obtained.

4.2. Variable Selection

The dependent variable, urban–rural income disparity, is operationalized as the urban-to-rural per capita income ratio (urban residents’ disposable income/rural residents’ disposable income). Elevated values denote amplified income differentials between urban and rural populations.

The core independent variable in this study is the Urban–Rural Construction Land Linkage Policy, represented as a quasi-natural experiment dummy variable. Municipalities implementing the policy are coded as 1 (treatment group), with non-implementing counterparts designated 0 (control group). Additionally, a time dummy variable is introduced to distinguish between the periods before and after policy implementation. The difference-in-differences (DIDs) interaction term, which is the product of the treatment group variable and the time dummy variable, serves as the key variable to assess the impact of the policy on urban–rural income disparity.

There are two mediating variables. For both mediating variables, we measure them in two ways for comparative analysis. Urbanization is not only a key driver of economic development but also plays a crucial role in promoting social structural transformation. Therefore, urbanization is a key transmission channel through which policy affects income inequality. It increases labor mobility and job opportunities, thereby potentially narrowing the income gap between urban and rural areas. The optimization and upgrading of industrial structure are critical indicators of a country’s economic maturity and efficiency. Specifically, upgrading the industrial structure, especially the expansion of the tertiary sector, tends to create higher-value jobs and reduce income disparities by creating employment and promoting economic inclusion. By continuously adjusting and improving the industrial structure, resources can be allocated more efficiently, production efficiency can be enhanced, and sustainable economic growth can be promoted.

This study considers the urbanization rate and industrial structure upgrading as the core mediating variables. It explores how the policy alleviates urban–rural income disparities by enhancing urbanization levels and industrial structure. Urbanization helps narrow the urban–rural income gap by absorbing rural labor into urban employment, facilitating industrial transformation, and revitalizing land resources. Industrial structure upgrading promotes sustained economic growth, creates job opportunities, and enhances labor productivity [50,51], contributing to a reduction in urban–rural income disparities. Following the research of Kuznets (1973) [52] and Xu (2015) [53], industrial structure upgrading is defined as follows:

$$str={\sum }_1^{3}i\times y_i=1\times y_1+2\times y_2+3\times y_3$$

(1)

$y_i$ is the proportion of $i$ industry to GDP, the range of industrial structure advanced is between 1 and 3, the closer it is to 1, the more the type of industry in the region is dominated by the primary industry and the level of industrial structure is low; if the index approaches a value of 2, this indicates that the regional industrial structure is predominantly characterized by secondary sector dominance, corresponding to an intermediate level of structural advancement. Conversely, an index nearing 3 signifies tertiary industry predominance within the regional economy, reflecting an advanced stage of industrial structure development.

Additionally, this study incorporates multiple control variables while also controlling for year (year) and region (id). These variables help account for other factors that may influence urban–rural income disparity. This allows for a more precise evaluation of the actual impact of the Urban–Rural Construction Land Increase–Decrease Linkage Policy on income disparity. Following the approaches of Wang and Li (2022) [54], Wang (2023) [55], Seven (2022) [56], Thornton and Tommaso (2020) [57], and Florian et al. (2022) [58], the selected control variables include the following:

• Regional economic development level, represented by the logarithm of per capita GDP;
• Local government fiscal self-sufficiency, measured by the ratio of local general budgetary revenue to local general budgetary expenditure. A higher value indicates greater fiscal independence;
• Regional financial development level, represented by the logarithm of year-end financial institution deposit balances;
• Inflation level, represented by the consumer price index (CPI);
• Population density, measured as the ratio of population to the administrative area of a region;
• The degree of openness, represented by total exports and imports as a share of GDP;
• Government behavior, represented by government expenditure as a share of GDP.

These control variables are selected to account for a broad range of socioeconomic and regional factors that may independently influence urban–rural income disparities. By controlling for differences in economic development level, fiscal capacity, financial infrastructure, inflation, demographic characteristics, and external openness, the model effectively isolates the specific effect of the Urban–Rural Construction Land Increase–Decrease Linkage Policy. This ensures that the observed impact on income inequality is not confounded by omitted variable bias, thereby enhancing the internal validity of the empirical findings.

Specific variables and descriptive statistics are presented in

Table 1. Variable descriptions.

Type	Variable Name	Code	Definition
Dependent Variable	Urban–Rural Income Disparity	incomegap	Urban residents’ per capita income/Rural residents’ per capita income
Independent Variable	Group	treat	Prefecture-level cities that implemented the revenue adjustment policy = 1; others = 0
	Policy Implementation	post	1 for years after policy implementation; 0 for years before policy implementation
	Difference-in-Differences	did	Interaction term: treat × post
Mediating Variable	Urbanization Rate [59]	city	(Urban population/Total population) × 100
		City1	The share of non-farm employment
	Industrial Structure	TP	Measured following Kuznets (1973) and Xu Min (2015) [52,53]
		TP1	Share of added value of tertiary industry in GDP [60]
Control Variable	Regional Economic Development Level [55]	pgdp	ln(per capita GDP)
	Local Government Fiscal Self-sufficiency [55]	selffin	Local government’s general fiscal revenue/general fiscal expenditure
	Regional Financial Development Level [56]	lfinlevl	ln(end-of-year deposits of financial institutions)
	Inflation Level [56]	cpi	Consumer Price Index
	Population Density [55]	popd	Population/Administrative area
	The degree of openness [57,59]	open	Total exports and imports as a share of GDP
	Government behavior [56,57]	gov	Government expenditure/GDP

and

Table 2. Descriptive statistics.

Code	Mean	Standard Deviation	Minimum	Median	Maximum
incomegap	2.372	0.731	1.207	2.281	26.301
did	0.032	0.176	0.000	0.000	1.000
pgdp	10.731	0.592	9.256	10.716	12.073
selffin	44.469	22.251	6.807	41.202	99.772
lfinlevel	16.883	1.084	14.894	16.748	20.156
cpi	102.251	1.203	99.130	102.100	105.714
popd	426.469	300.904	18.000	370.000	1440.000
open	681.746	2096.953	0.230	192.360	17927.355
gov	0.214	0.147	0.075	0.176	1.159

. All actual values are deflated using the GDP deflator, with 2010 as the base year.

Variable definitions and measurement details are summarized in Table 1. Table 2 reports the descriptive statistics for all variables.

4.3. Empirical Methods

4.3.1. Benchmark Regression

Since the policy of “urban–rural construction land transfer” is implemented in phases and regions, this paper uses a multi-period DID model to examine the specific impact of the policy on the urban–rural income gap. The cities where the policy is implemented are the treatment group, while other cities are the control group. According to the definition of the difference-in-differences (DID) model, a multi-period DID difference term is constructed. The basic DID model is specified as follows:

$${\mathrm{i}\mathrm{n}\mathrm{c}\mathrm{o}\mathrm{m}\mathrm{e}gap}_{\mathrm{i}\mathrm{t}}={\alpha }_0+{\alpha }_1{\mathrm{D}\mathrm{I}\mathrm{D}}_{\mathrm{i}\mathrm{t}}+{\alpha }_2\sum {\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}\mathrm{s}}_{\mathrm{i}\mathrm{t}}+{\delta }_{\mathrm{t}}+{\mu }_{\mathrm{i}}+{\epsilon }_{\mathrm{i}\mathrm{t}}$$

(2)

In the above model, i represents the city, t represents time, ${\mathrm{i}\mathrm{n}\mathrm{c}\mathrm{o}\mathrm{m}\mathrm{e}gap}_{\mathrm{i}\mathrm{t}}$ is the dependent variable, ${\alpha }_0$ is the constant term, ${\alpha }_1$ and ${\alpha }_2$ are the regression coefficients, ${\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}\mathrm{s}}_{\mathrm{i}\mathrm{t}}$ represents control variables, and ${\epsilon }_{\mathrm{i}\mathrm{t}}$ is the random error term. Furthermore, this paper introduces fixed effects for both cities and time. ${\delta }_{\mathrm{t}}$ is used to measure the time effects, while ${\mu }_{\mathrm{i}}$ measures the fixed effects at the individual level. By simultaneously controlling for both time and individual fixed effects, the model can effectively address the issue of omitted variables that vary across individuals but not over time or that vary over time but not across individuals.

4.3.2. Mechanism Effect Model

This investigation employs a mechanism effect model to examine the operational pathways through which the Urban–Rural Construction Land Linkage Policy influences the urban–rural income disparity. The analytical procedure comprises three sequential phases: initially establishing the regression framework through Equations (3) and (4), followed by separate assessment of policy effects on dual mediating variables utilizing Equation (3). Ultimately, Equation (4) is implemented to verify the two mediating variables by which the policy affects income inequality between urban and rural sectors.

$${\mathrm{Y}}_{\mathrm{i}\mathrm{t}}={\beta }_0+{\beta }_1{\mathrm{D}\mathrm{I}\mathrm{D}}_{\mathrm{i}\mathrm{t}}+{\beta }_2\sum {\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}\mathrm{s}}_{\mathrm{i}\mathrm{t}}+{\delta }_{\mathrm{t}}+{\mu }_{\mathrm{i}}+{\epsilon }_{\mathrm{i}\mathrm{t}}$$

(3)

$${\mathrm{i}\mathrm{n}\mathrm{c}\mathrm{o}\mathrm{m}\mathrm{e}gap}_{\mathrm{i}\mathrm{t}}={\gamma }_0+{\gamma }_1{\mathrm{Y}}_{\mathrm{i}\mathrm{t}}+{\gamma }_2{\mathrm{D}\mathrm{I}\mathrm{D}}_{\mathrm{i}\mathrm{t}}+{\gamma }_3\sum {\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}\mathrm{s}}_{\mathrm{i}\mathrm{t}}+{\delta }_{\mathrm{t}}+{\mu }_{\mathrm{i}}+{\epsilon }_{\mathrm{i}\mathrm{t}}$$

(4)

where ${\mathrm{Y}}_{\mathrm{i}\mathrm{t}}$ is the mediating variables, parameters ${\beta }_0$ and ${\gamma }_0$ denote constant terms, and ${\gamma }_1$ represents the regression coefficient for the mediating variable. The components ${\beta }_1$ and ${\gamma }_2$ constitute the key estimable parameters, with ${\epsilon }_{\mathrm{i}\mathrm{t}}$ specified as a normally distributed stochastic error term.

5. Results

5.1. Baseline Model Regression Results

Table 3. Result of baseline regression.

	(1)	(2)	(3)	(4)
Variable	Incomegap	Incomegap	Lruralincome	Lurbanincome
did	−0.156 **	−0.107 *	0.062 ***	0.043 ***
	(0.061)	(0.056)	(0.017)	(0.016)
pgdp		−0.113	0.150 ***	0.081 ***
		(0.127)	(0.041)	(0.027)
selffin		−0.002	0.001 *	0.001 *
		(0.002)	(0.001)	(0.000)
lfinlevel		−0.406	0.059	−0.024
		(0.292)	(0.065)	(0.043)
cpi		−0.022	−0.007	−0.010 **
		(0.023)	(0.004)	(0.004)
popd		−0.000	−0.000	−0.000
		(0.000)	(0.000)	(0.000)
open		0.000	−0.000	−0.000
		(0.000)	(0.000)	(0.000)
gov		−1.072	0.301 *	0.079
		(0.726)	(0.158)	(0.053)
Constant	2.377 ***	13.051 *	7.410 ***	10.796 ***
	(0.002)	(6.458)	(1.223)	(0.907)
Observations	3822	3783	3783	3783
R-squared	0.495	0.505	0.961	0.962
Individual	Yes	Yes	Yes	Yes
Time	Yes	Yes	Yes	Yes
Adj R-squared	0.451	0.460	0.957	0.959

Note: *** p < 0.01, ** p < 0.05, * p < 0.1, standard deviation in parentheses. Source: Authors’ calculations.

reports the baseline regression results of how the Urban–Rural Construction Land Linkage Policy affects urban–rural income disparity. Column (1) shows the results without including control variables, column (2) shows the results with control variables included, and columns (3) and (4) show the regression results for rural residents’ per capita income and urban residents’ income, respectively. One can obtain the following information from the table. As presented in column (1), the DID estimator yields a coefficient of −0.156 (p < 0.05) without control variables, demonstrating the policy’s significant narrowing effect on urban–rural income disparities. Column (2) shows that with the inclusion of control variables, the policy effect is −0.107, still significant at the 10% level of significance, thus statistically verifying hypothesis H1. Notably, the land transfer policy implementation exhibits asymmetric income effects: while increasing per capita income in both urban and rural areas, its marginal impact proves stronger on rural residents’ earnings through rural productivity enhancement channels.

5.2. Robustness Test

To strengthen the analytical rigor, we performed robustness verification through three methodological approaches: parallel trend diagnostics, placebo examination, and propensity score matching difference-in-differences (PSM-DID) estimation.

5.2.1. Parallel Trend Test

The parallel trend assumption is a crucial prerequisite for the validity of the difference-in-differences (DID) approach. This assumption posits that, in the absence of treatment, the treatment and control groups would have followed similar trajectories over time. Violation of this condition implies that the estimated treatment effect from the DID approach may not accurately capture the true causal effect of the policy; instead, it could reflect inherent, pre-existing differences between the two groups. To empirically verify this assumption, this study employs an event–study framework. Specifically, an interaction term is constructed by multiplying the time-period dummy variables (time) for each period with the treatment-group dummy variable (treat). These interaction terms are subsequently included in the regression to capture any deviations in trends between treatment and control groups over time.

Figure 2 graphically presents the results of the parallel trend analysis. The pre-implementation phase coefficients (pre_2 to pre_4) show non-significant estimates with confidence intervals encompassing zero, confirming comparable pretreatment trends between treatment and control cohorts. Conversely, the coefficients for periods following policy implementation (post_2 to post_4) are statistically significant, as indicated by confidence intervals that exclude zero. This implies significant divergence in outcomes between the two groups after policy introduction. These temporal dynamics collectively confirm the satisfaction of parallel trend assumptions, substantiating the DID framework’s internal validity.

5.2.2. Placebo Test

To address endogeneity concerns arising from unobserved omitted variables in the baseline specification, the analysis implements statistical placebo testing. Specifically, a fictitious policy timeline is constructed, and placebo cities are selected to substitute the original control group cities. Subsequently, the urban–rural income gap is re-estimated for the hypothetical treatment group through 500 repeated regressions. Figure 3 presents the probability density distributions of the resulting coefficient estimates and their associated p-values. As illustrated, most of these placebo estimates cluster around zero, conforming to a normal distribution and demonstrating statistical insignificance. Importantly, the actual baseline regression coefficient estimate is located to the left of the distribution obtained from placebo regressions. This result significantly diminishes concerns that the observed effects in the baseline estimates are attributable to unobserved factors, thus supporting the robustness of the findings presented in this paper.

5.2.3. PSM-DID Model

To further improve the credibility of the DID results, this investigation implements the propensity score matching (PSM) methodology to address inter-group heterogeneity. The DID framework fundamentally requires compliance with the parallel trends’ assumption—namely, that treatment and control cohorts exhibit statistically indistinguishable pretreatment trajectories. This identification condition ensures counterfactual equivalence between groups prior to policy intervention, thereby mitigating selection bias and endogenous interference. However, due to individual differences, the samples may exhibit different characteristics and trends, which could interfere with the random assignment of the sample to either the treatment or control group, leading to biased estimates. To tackle this issue, the study employs the Propensity Score Matching-Difference in Differences (PSM-DID) approach. Prior to DID implementation, untreated cities sharing similar characteristics with the treatment group are systematically selected as counterfactual matches. This dual-stage methodology effectively reduces sample selection bias while ensuring temporal trend comparability between experimental and control groups.

This paper uses a one-to-two logit model matching method for PSM matching. After matching propensity scores through logit regression, tests can be conducted. First, a common support test is performed. If the propensity scores of the two groups are close, it indicates that the common support assumption is satisfied, followed by a balance test for the two groups. This study employs nearest neighbor matching to pair treatment and control groups. As demonstrated in

Table 4. PSM matching result.

	Unmatched	Mean			%reduct	t-Test
Variable	Matched	Treated	Control	%bias	bias	t	p > t
selffin	U	30.423	45.216	−76.4		−6.58	0
	M	30.423	30.505	−0.4	99.4	−0.04	0.964
lfinlevel	U	17.242	16.873	40.3		3.62	0
	M	17.242	17.245	−0.3	99.2	−0.03	0.979
popd	U	406.91	434.4	−9.6		−0.90	0.371
	M	406.91	388.64	6.4	33.6	0.48	0.630

, the propensity scores matching balance tests reveal that post-matching mean measurements between the treatment and control cohorts exhibit minimal differences. The results show that, compared to the unmatched sample, the standardized bias of all covariates was reduced after matching in the matched sample (Matched), with absolute values close to 10%. Additionally, the p-values of the observable variables after matching are all greater than 0.1, indicating a good matching effect.

Figure 4 shows the balance test results of the matching process. The test graph indicates that, before matching, the standardized bias of the variables is large, and the standard deviation values of the variables are high. After matching, the standardized bias is significantly reduced. Therefore, if the two groups are regressed directly without matching, severe estimation errors may arise. By removing the unmatched samples, the remaining cities in the matched sample are very close in terms of key characteristic variables, and the selection bias is significantly reduced, meeting the comparability requirements. The systematic differences between untreated and treated individuals are noticeably decreased, and the matching effect is good.

Figure 5 shows the propensity score density functions before and after matching. The common support assumption requires that the distribution of the kernel density plots be roughly the same after matching. This is verified by comparing the propensity score (Pscore) values before and after matching. Observing the left graph, the distribution center of the control group’s propensity scores is noticeably higher than that of the experimental group before matching. After matching, the control group’s distribution center is clearly closer to the experimental group and has become consistent, with the distance between the centers shortened. This indicates that PSM has effectively corrected the sample selection bias, and the matching results are ideal, satisfying the common support assumption.

Based on the matching results, the study eliminates unmatched observations, with

Table 5. PSM-DID model results.

	(1)	(2)	(3)	(4)
Variable	Incomegap	Incomegap	Ruralincome	Urbanincome
did	−0.109 **	−0.090 **	0.055 ***	0.036 **
	(0.042)	(0.041)	(0.014)	(0.016)
pgdp		−0.078	0.105 ***	0.087 **
		(0.062)	(0.025)	(0.034)
selffin		0.001	0.002 *	0.002 **
		(0.001)	(0.001)	(0.001)
lfinlevel		−0.067	−0.020	−0.059
		(0.075)	(0.039)	(0.037)
cpi		−0.005 *	0.001	−0.001
		(0.003)	(0.001)	(0.001)
popd		0.001 *	−0.000 **	−0.000 **
		(0.000)	(0.000)	(0.000)
open		0.000	−0.000	0.000
		(0.000)	(0.000)	(0.000)
gov		−0.018 **	0.006 **	0.001
		(0.009)	(0.003)	(0.003)
Constant	2.395 ***	4.529 ***	8.674 ***	10.390 ***
	(0.002)	(1.522)	(0.858)	(0.863)
Observations	2782	2756	2756	2756
R-squared	0.899	0.906	0.986	0.981
Individual	Yes	Yes	Yes	Yes
Province	Yes	Yes	Yes	Yes
Time	Yes	Yes	Yes	Yes
Adj R-squared	0.887	0.895	0.984	0.979

Note: *** p < 0.01, ** p < 0.05, * p < 0.1, standard deviation in parentheses. Source: Authors’ calculations.

displaying the DID regression outcomes following PSM adjustment. The DID variable coefficients measure −0.109 and −0.090, both attaining statistical significance at the 5% level. Relative to baseline model estimates, the PSM-adjusted specification maintains statistical robustness, confirming that land transfer policy enactment significantly diminishes urban–rural income disparity.

5.3. Heterogeneity Analysis

Considering China’s vast territory and significant regional disparities, different regions exhibit distinct natural conditions, economic development levels, and cultural backgrounds, leading to uneven regional development outcomes. Therefore, the same policy may yield heterogeneous effects across regions. To further examine the generalizability of the findings, this study conducts heterogeneity analysis through stratified data processing. First, cities are categorized into eastern, central, and western regional subsamples for separate regression analyses. Subsequently, using 2016 as the reference year, cities are further stratified by the median economic development level into two subsamples, categorized as economically developed versus relatively underdeveloped, which are then analyzed separately through regression. The heterogeneity analysis outcomes are presented in

Table 6. Heterogeneity analysis results.

	(1)	(2)	(3)	(4)	(5)
Variable	Incomegap	Incomegap	Incomegap	Incomegap	Incomegap
did	0.061	−0.077 *	−0.169 *	−0.062 ***	−0.067
	(0.048)	(0.031)	(0.092)	(0.024)	(0.210)
pgdp	−0.079 *	−0.252 *	0.221	−0.188 ***	0.170
	(0.042)	(0.107)	(0.303)	(0.026)	(0.121)
selffin	−0.000	−0.001	0.003 ***	−0.002 ***	−0.003
	(0.000)	(0.001)	(0.001)	(0.001)	(0.003)
lfinlevel	−0.011	0.021	−0.793	−0.030	−0.795 ***
	(0.045)	(0.088)	(0.753)	(0.028)	(0.168)
cpi	0.017	−0.013	−0.012	−0.013 **	−0.024
	(0.011)	(0.011)	(0.010)	(0.006)	(0.032)
popd	0.001 ***	−0.000	−0.001 **	−0.001 ***	−0.000
	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)
open	−0.000	0.000	0.000	0.000***	−0.000
	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)
gov	−0.622 ***	−0.619	0.022	−0.189***	−1.802 ***
	(0.116)	(0.596)	(0.031)	(0.060)	(0.314)
Constant	0.988	6.219 **	14.878	6.553 ***	17.160 ***
	(1.366)	(2.043)	(10.031)	(0.801)	(4.294)
Observations	1131	1040	1170	1872	1911
R-squared	0.804	0.908	0.322	0.926	0.324
Individual	Yes	Yes	Yes	Yes	Yes
Time	Yes	Yes	Yes	Yes	Yes
Adj R-squared	0.783	0.898	0.252	0.919	0.260

Note: *** p < 0.01, ** p < 0.05, * p < 0.1, standard deviation in parentheses. Source: Authors’ calculations.

.

Column (1) displays regression outcomes for eastern China’s subsample. The land transfer policy exhibits no statistically significant impact on urban–rural income disparities in this region. This phenomenon may stem from eastern China’s indistinct urban–rural demarcation (e.g., Pearl River Delta and Yangtze River Delta), where mature rural land markets coexist with constrained reclamation potential from rural construction lands, collectively curtailing policy efficacy in resource optimization. Additionally, the region’s pre-existing narrow urban–rural income differential provides limited potential for substantial reduction through policy intervention.

Columns (2) and (3), respectively, present regression results for central and western regional subsamples. The DID coefficient in column (2) stands at −0.077 with 10% statistical significance, while column (3) shows a stronger coefficient of −0.169 significant at the 10% level. Comparative coefficient magnitudes reveal intensified policy effectiveness in western regions, attributable to their more extensive rural construction land reserves, enhanced reclamation capacity, and heightened responsiveness to land quota trading mechanisms. This pattern confirms the policy’s superior efficacy in reducing urban–rural income disparities in western China.

Columns (4) and (5) present findings for economically underdeveloped and relatively developed regions, respectively. The analysis reveals that the land increase–decrease Linkage Policy significantly reduces the urban–rural income gap in underdeveloped regions (significant at the 1% level), while the policy effect remains statistically insignificant in developed regions. This spatial correspondence aligns with China’s regional development pattern, where economically advanced cities predominantly cluster in eastern regions, while underdeveloped cities concentrate in western areas. So, this result is similar to the result of the division of the east, middle, and west in the previous article. Collectively, these results demonstrate pronounced policy effectiveness in western regions and economically disadvantaged areas.

China’s regional division into eastern, central, and western regions primarily reflects economic and geographic distinctions. Furthermore, the analysis adheres to China’s statutory six-region classification system encompassing north China, northeast China, east China, south–central China, southwest China, and northwest China—a comprehensive spatial division established by the State Council that integrates natural geography, historical evolution, and ethnic distribution characteristics. The regression results under this state-mandated regional classification are presented in Table 1. (There are no cities in northeast China that implement the policy, so there are only five regions in the results).

Table 7. Another division result.

	(North China)	(East China)	(South Central)	(Southwest)	(Northwest)
Variable	Incomegap	Incomegap	Incomegap	Incomegap	Incomegap
did	−0.112 **	0.048	−0.136 ***	−0.090 ***	0.284
	(0.049)	(0.037)	(0.044)	(0.028)	(0.961)
pgdp	−0.131 **	−0.169 ***	−0.373 ***	−0.085 ***	0.641
	(0.054)	(0.034)	(0.063)	(0.027)	(0.667)
selffin	0.000	−0.003 ***	0.000	0.003 ***	0.011
	(0.001)	(0.001)	(0.001)	(0.001)	(0.013)
lfinlevel	−0.067	−0.068	−0.012	−0.080 ***	−2.974 ***
	(0.066)	(0.047)	(0.064)	(0.030)	(0.759)
cpi	0.009	0.030 ***	−0.032 **	−0.003	−0.117
	(0.009)	(0.010)	(0.013)	(0.008)	(0.124)
popd	0.000	−0.000	−0.000	−0.000	0.001
	(0.000)	(0.000)	(0.000)	(0.000)	(0.004)
open	−0.000 ***	−0.000	0.000***	−0.000	0.000
	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)
gov	−0.764 ***	−0.601 ***	−1.648 ***	−0.003	−2.873 **
	(0.290)	(0.122)	(0.359)	(0.047)	(1.364)
Constant	4.206 ***	2.499 *	9.967 ***	5.155 ***	56.684 ***
	(1.455)	(1.294)	(1.794)	(0.950)	(16.500)
Observations	429	1001	1040	481	390
R-squared	0.934	0.870	0.849	0.953	0.310
Individual	Yes	Yes	Yes	Yes	Yes
Time	Yes	Yes	Yes	Yes	Yes
Adj R-squared	0.925	0.857	0.833	0.947	0.210

Note: *** p < 0.01, ** p < 0.05, * p < 0.1, standard deviation in parentheses. Source: Authors’ calculations.

reveals the differentiated spatial effectiveness of the land policy: statistically insignificant impacts in east China and northwest China. This pattern aligns with the mechanisms observed in eastern regions discussed earlier. In East China, characterized by the most advanced regional economy, advanced marketization, and blurred urban–rural boundaries, the policy demonstrates limited effectiveness. Northwest China’s insignificant policy outcomes likely stem from challenging topography/climate conditions (inferior resource endowment), coupled with weaker administrative implementation capacity. Conversely, the policy achieves significant urban–rural income gap reduction in central–south, southwest, and north China, particularly in the central–south, where China’s primary arable land reserves concentrate. These geographically differentiated outcomes correspond to China’s regional development realities.

5.4. Dynamic Effect

To capture the potential dynamic response paths of the policy, this paper follows the approaches of Furceri et al. (2016) [61] and Malla et al. (2022) [62] by constructing one-period lag (did_lag1) and two-period lag (did_lag2) of the policy variable to examine its persistent effects. The lagged variables are dynamically generated based on the implementation year of city-level policies, ensuring that non-policy intervention periods are not erroneously incorporated into the policy response paths, thereby avoiding the issue of “future information backflow”. As shown in

Table 8. Dynamic analysis results.

	(1)	(2)	(3)	(4)
Variable	Incomegap	Incomegap	Lruralincome	Lurbanincome
did	−0.130 **	−0.098 **	0.059 ***	0.037 **
	(0.051)	(0.047)	(0.016)	(0.014)
did_lag1	−0.023	−0.023	0.006	0.003
	(0.020)	(0.026)	(0.005)	(0.003)
did_lag2	−0.012	−0.006	0.012	0.012
	(0.029)	(0.032)	(0.009)	(0.009)
pgdp		0.004	0.100 *	0.064 **
		(0.179)	(0.053)	(0.029)
selffin		0.000	0.000	0.000
		(0.001)	(0.000)	(0.000)
lfinlevel		−0.327	0.027	−0.032
		(0.316)	(0.074)	(0.043)
cpi		−0.011	0.001	−0.001
		(0.008)	(0.001)	(0.002)
popd		0.001	−0.000	−0.000 **
		(0.000)	(0.000)	(0.000)
open		0.000	−0.000	−0.000
		(0.000)	(0.000)	(0.000)
gov		−0.026	0.011	−0.000
		(0.028)	(0.008)	(0.003)
Constant	2.377 ***	8.778 *	7.924 ***	10.268 ***
	(0.002)	(4.354)	(1.144)	(0.821)
Observations	3822	3783	3783	3783
R-squared	0.495	0.497	0.960	0.962
Individual	Yes	Yes	Yes	Yes
Province	Yes	Yes	Yes	Yes
Time	Yes	Yes	Yes	Yes
Adj R-squared	0.451	0.451	0.956	0.959

Note: *** p < 0.01, ** p < 0.05, * p < 0.1, standard deviation in parentheses. Source: Authors’ calculations.

, the regression results indicate that the policy significantly affects the urban–rural income gap and residents’ income levels in the current period, but its lagged effects are statistically insignificant. This suggests that the policy tends to generate immediate stimulus effects rather than long-term persistence, and subsequent institutional support may be required to extend its efficacy.

5.5. Mechanism Test

Regarding the mechanism test, Xia (2019) argued that the surge in fiscal revenue from cross-regional trading of the land transfer surplus indicators might disrupt the existing regional development model, potentially leading to the loss of endogenous growth momentum in the region [63]. Lu (2021) found that excessive reliance on land finance may lead to competition among local governments, hinder the process of regional integration, and exacerbate regional development imbalances [64]. Therefore, to test whether this policy could mislead local governments and cause impoverished regions to lose their endogenous development momentum, it is necessary to examine the policy’s effect on the endogenous mechanisms of regional economic development. The endogenous driving force of economic development comes from the processes of industrialization and urbanization [65]. This paper also takes these two perspectives into account, selecting industrial structure upgrading and urbanization rate as variables, and constructs a mechanism verification model. The empirical methodology incorporates Hausman test findings yielding a p-value of 0.0001 that categorically rejects the null hypothesis, necessitating selection of the fixed-effects specification.

5.5.1. Urbanization Rate

Additionally,

Table 9. Mechanism effect test results for the urbanization rate.

	(1)	(2)	(3)	(4)
Variable	City	Incomegap	City1	Incomegap
did	0.018 **	−0.116 *	−0.006	−0.119 *
	(0.007)	(0.070)	(0.010)	(0.070)
city		−0.278 *
		(0.167)
city1				0.216 *
				(0.121)
pgdp	−0.021 ***	−0.002	−0.013 *	0.007
	(0.005)	(0.049)	(0.007)	(0.049)
selffin	0.001 ***	0.000	0.000	0.000
	(0.000)	(0.001)	(0.000)	(0.001)
lfinlevel	0.079 ***	−0.305 ***	0.069 ***	−0.342 ***
	(0.008)	(0.075)	(0.010)	(0.074)
cpi	0.001	−0.011	−0.001	−0.011
	(0.001)	(0.007)	(0.001)	(0.007)
popd	−0.000	0.000 ***	−0.000 ***	0.001 ***
	(0.000)	(0.000)	(0.000)	(0.000)
open	−0.000	0.000	−0.000	0.000
	(0.000)	(0.000)	(0.000)	(0.000)
gov	0.004	−0.025	0.009 ***	−0.028
	(0.002)	(0.023)	(0.003)	(0.023)
Constant	−0.626 ***	8.610 ***	−0.366 *	8.864 ***
	(0.150)	(1.476)	(0.207)	(1.473)
Observations	3783	3783	3783	3783
R-squared	0.899	0.497	0.982	0.497
Individual	Yes	Yes	Yes	Yes
Province	Yes	Yes	Yes	Yes
Time	Yes	Yes	Yes	Yes
Adj R-squared	0.890	0.452	0.980	0.452

Note: *** p < 0.01, ** p < 0.05, * p < 0.1, standard deviation in parentheses. Source: Authors’ calculations.

presents the mechanism effect analysis for the urbanization rate. Columns (1) and (2) indicate that the land transfer policy significantly promotes urbanization, and this increase in urbanization significantly reduces the urban–rural income gap. After incorporating the urbanization rate into the baseline model, the coefficients on the estimates of DID and city are both negative and remain statistically significant, thus validating Hypothesis 2a. Columns (3) and (4) show the results of using “share of non-farm population” as a measure of the urbanization rate, which is unsatisfactory, and may be a more appropriate indicator to use “urban population” as a measure of the urbanization rate.

5.5.2. Upgrading of an Industrial Structure

Furthermore, this paper uses industrial structure upgrading as a proxy variable for industrial development, with regression outcomes presented in

Table 10. Mechanism effect test results for the industrial structure.

	(1)	(2)	(3)	(4)
Variable	TP	Incomegap	TP1	Incomegap
did	0.016 **	−0.112	0.258 *	−0.120 *
	(0.008)	(0.070)	(0.155)	(0.070)
TP		−0.565 ***
		(0.154)
TP1				−0.002
				(0.008)
pgdp	−0.012 **	−0.003	−1.392 ***	0.002
	(0.005)	(0.049)	(0.108)	(0.050)
selffin	−0.000 **	0.000	−0.001	0.000
	(0.000)	(0.001)	(0.002)	(0.001)
lfinlevel	0.023 ***	−0.314 ***	1.722 ***	−0.324 ***
	(0.008)	(0.074)	(0.164)	(0.075)
cpi	0.001	−0.011	0.020	−0.011
	(0.001)	(0.007)	(0.015)	(0.007)
popd	−0.000	0.000 ***	−0.000	0.001 ***
	(0.000)	(0.000)	(0.000)	(0.000)
open	0.000	0.000	−0.000	0.000
	(0.000)	(0.000)	(0.000)	(0.000)
gov	0.049 ***	0.002	3.095 ***	−0.020
	(0.002)	(0.024)	(0.050)	(0.033)
Constant	1.988 ***	9.908 ***	−16.150 ***	8.755 ***
	(0.162)	(1.501)	(3.269)	(1.478)
Observations	3783	3783	3783	3783
R-squared	0.856	0.499	0.684	0.497
Individual	Yes	Yes	Yes	Yes
Province	Yes	Yes	Yes	Yes
Time	Yes	Yes	Yes	Yes
Adj R-squared	0.844	0.454	0.656	0.452

Note: *** p < 0.01, ** p < 0.05, * p < 0.1, standard deviation in parentheses. Source: Authors’ calculations.

. Columns (1) and (2) demonstrate that the DID variable significantly promotes industrial structure upgrading at the 5% confidence level. Incorporating industrial structure upgrading into the baseline model reveals that industrial upgrading significantly narrows the urban–rural income gap, suggesting that the land transfer policy reduces income disparity by facilitating industrial optimization and upgrading, thereby validating Hypothesis 2b. After replacing the measure of industrial structure upgrading with TP1 in columns (3) and (4), the results still indicate that the land policy promotes industrial structure upgrading, and thus contributes to the reduction in the urban–rural income gap.

The mechanistic investigation elucidates dual pathways through which the land transfer policy attenuates urban–rural income disparities: facilitating industrial structural advancement and accelerating urbanization processes. This regulatory framework streamlines land resource allocation, stimulates non-agricultural labor relocation, and orients capital influx toward rural sectors. These coordinated effects collectively propel industrial composition refinement, elevate urbanization indices, and amplify rural household earnings, culminating in the attenuation of inter-sectoral income inequality. The operational architecture of these causal pathways is schematically represented in Figure 6.

6. Conclusions and Recommendations

This research examines the impact of the Urban–Rural Construction Land Linkage Policy on income disparities through a multi-period DID framework. Municipalities implementing the policy constitute the treatment cohort, while non-implementing cities form the control group. The analysis assesses regional variations in urban–rural income gaps across pre- and post-policy implementation periods. Following robustness verification and mechanistic investigation, the core findings demonstrate the following:

The policy has significantly narrowed the urban–rural income gap in the treated regions, affirming the effectiveness of land transfer mechanisms in alleviating regional income inequality. Model robustness is further confirmed through parallel trend tests, placebo tests, and PSM-DID estimations.

Heterogeneity analysis reveals that policy effects vary across regions. In the eastern region and high-income areas, the policy’s impact is statistically insignificant, likely due to blurred urban–rural boundaries and mature rural land markets with limited room for land reclamation. In contrast, the policy demonstrates stronger effects in western regions, where abundant rural construction land and greater land restoration capacity amplify the effectiveness of land quota trading in narrowing income gaps.

Regarding temporal dynamics, the policy significantly reduces income disparities in the implementation year, but its effects weaken and become statistically insignificant after one- and two-year lags. Mechanism analysis reveals that the policy exerts its effects primarily through two channels: promoting urbanization and facilitating industrial structure upgrading.

Since the implementation of the Linkage Policy, the income gap between urban and rural residents in the areas where the policy is implemented has narrowed significantly, but there are still some deficiencies in the implementation of the policy. According to the above research conclusions, we propose the following suggestions:

(1)

Protecting the rights and interests of farmers and improving the policy system. There are always some problems in the implementation of the policy of “linking the increase and reduction in urban and rural construction land”, for example, the urban construction land is mostly occupied by high-quality arable land, while the rural arable land supplemented by collation is poor and low-quality arable land, and the local government will be guided by the increase in the scale of the urban construction land, which will lead to the demolition of the old areas without the principle of the rules and regulations, and the abuse of the right of land acquisition. The right of land expropriation. Therefore, it is necessary to continuously improve the policy and system to effectively protect farmers’ land rights and interests and to ensure that rural residents enjoy the dividends of land appreciation brought about by urbanization;

(2)

Reasonable land use arrangement and optimization of industrial structure. To a certain extent, the Linkage Policy promotes the upgrading of industrial structure and the growth of farmers’ income; therefore, the implementation of the policy needs to take the project of linkage between urban and rural construction land, optimize the industrial structure and spatial layout, and reasonably arrange the agricultural land and construction land with the core of saving and intensive land use, using the projects of linkage between urban and rural construction land, and so on, as a carrier;

(3)

Promote market-oriented reform and improve policy efficiency. The role of the market in the allocation of land resources has been increasing, and the market has been used to revitalize the stock of land and low-efficiency land, achieve the optimal allocation of land elements, and promote the balanced development of the economy in various regions;

(4)

Increasing long-term supporting measures. The empirical results show that the Linkage Policy has a significant effect in promoting income increase and income disparity reduction in the short term, but the impact is insufficient in the long term. Therefore, it is recommended that the long-term implementation mechanism of the policy be further improved and the continuity and stability of the policy be strengthened. For example, by establishing a dynamic evaluation mechanism, improving the follow-up input system of financial funds, and promoting the assessment of policy objectives and responsibilities, it is recommended to ensure that the policy effect can be transformed from a short-term stimulus to a medium- and long-term institutional dividend.